Yardstick Research tear-sheet / healthcare RCM cohort

Methodology · how we score · rubric weights in plain sight · vendors received this sheet seven days before publication and could flag factual errors, never rankings

Nym-Health

Identity

Nym (often "Nym Health"). Autonomous medical coding for hospitals, health systems, and physician groups. Founded 2018, Tel Aviv R&D + New York HQ. Founders Amihai Neiderman and Adam Rimon (both Israel Defense Forces Unit 8200 alums); Lori M. Jones appointed CEO April 21, 2026 (succeeded Or Peles, 2022-2026). $94.5M total funding across four rounds (Bessemer Seed 2019, GV Series A 2020, Addition Series B 2021, PSG Series C $47M Oct 2024). ~290 employees per LinkedIn / CB Insights Q2 2026. Six supported specialties: emergency medicine, radiology, outpatient surgery, outpatient visits, inpatient professional services, urgent care. Rules-based "Clinical Language Understanding" (CLU) architecture - deliberately non-LLM, audit-trail-first.

Total score: 66 / 100

Computed: 71.25 weighted dim total minus 5 for soft pricing-transparency penalty = 66.25 → 66.

• Stage fit:
• Foundation (<40 readiness): no - enterprise sales motion, no self-service, and the architecture is built for production-scale ED + radiology coding rather than discovery-stage evaluation.
• Pilot (40-59): conditional - Nym does run phased rollouts (single-specialty + single-facility pilots ramping to enterprise), and KLAS customers specifically rated implementation + ongoing support as high-satisfaction. But pricing is quote-only, so a pilot still requires entering an enterprise sales cycle.
• Scale (60-79): yes - this is the heartland. Geisinger ED (20 minutes → 2.5 seconds), Genesis ED (50% → 70% automation at 96%+ accuracy), Inova ED multi-year customer with named HIM AVP on the record. 6M+ charts annually across 400+ facilities and 21 named providers per the Series C announcement.
• Optimization (80+): conditional - strong on outpatient + inpatient-professional optimization. Wrong-fit if the optimization scope includes DRG coding (out of Nym scope; need Iodine / AKASA / CodaMetrix for that surface).
• One-line verdict: The rules-based audit-trail counter-pattern in autonomous coding - KLAS Top Performer at integrated-delivery-network scale, strongest on ED and radiology, materially weaker security-posture story than HITRUST-certified peers.

Headline numbers

Dimension scores

Subtotal weighted dims: 71.25 Pricing-transparency penalty: -5 (soft, enterprise column) Total: 66.25 → 66

Pricing detail

• Public price tiers: None. /pricing returns HTTP 404. No per-chart, per-encounter, per-coder, or per-specialty anchor published on any vendor surface.
• Outcome anchors (vendor claims, "up to" framing): Cost reductions up to 35%; revenue capture lift up to 5%; claim denial reduction up to 50%; AR-days improvement up to 5 days.
• Pricing model (assumed by sector convention, not Nym-confirmed): Per-chart or per-encounter with volume discounts; share-of-savings or hybrid is also plausible. Cannot verify without entering a sales cycle.
• Pricing-transparency penalty: soft (enterprise column, since size_fit_min = $250M).

Integrations

• Epic: Native - Epic Toolbox listed (Fully Autonomous Coding category, effective Aug 16, 2024).
• athenahealth: Compatibility claimed (MedCity News Oct 2024); no athenaConnect marketplace listing surfaced as of May 2026.
• Oracle-Cerner: Compatibility claimed; no named customer or marketplace listing.
• Meditech: Compatibility claimed; no named customer reference.
• Allscripts: Compatibility claimed; no named customer reference.
• FHIR R4: First-party strategy - "eliminates extensive pre-configuration" for FHIR-supporting clients per Nym's integration blog. Director of Integration is ex-Epic (Austin McCrimmon).
• HL7 v2: Implicit (Epic Toolbox listing mechanics; non-FHIR-mature EHRs).
• SMART on FHIR: Not claimed.
• Native cohort-peer integrations: None documented in current Yardstick D1 cohorts.

Editorial assessment

Nym is the deliberate counter-pattern in the autonomous-coding cohort. While AKASA, Fathom, CodaMetrix, and SmarterDx have all moved toward deep-learning or LLM-based architectures, Nym has stayed on a rules-based "Clinical Language Understanding" path and explicitly markets the audit-trail story as the differentiator. For HIM directors who need to defend every code assignment to a payer or auditor - and for emergency-department leaders where the chart documentation is structured enough that rules-based extraction works well - that architectural choice is genuinely buyer-relevant. The KLAS Top Performer status (89.6/100 in the first-ever KLAS autonomous-coding report), Epic Toolbox listing, named-customer density (Geisinger, Ochsner, Inova, Henry Ford, Loma Linda, Genesis, OSF), and case-study outcome metrics (Geisinger ED 20 minutes → 2.5 seconds, denials under 0.1%; Genesis ED 50% → 70% automation at 96%+ accuracy) hold up under buyer scrutiny.

The bounded weak spots map cleanly. The security-posture gap is real: "mapped to HiTrust" is a meaningfully weaker claim than the HITRUST CSF Certification that AKASA (r2), Cohere (r2), Fathom (i1), and Hippocratic (e1) all carry. The absence of a public trust page (/trust and /security both return HTTP 404) is the procurement-friction expression of the same issue. The leadership-continuity signal warrants attention: three CEOs in eight years, with the April 2026 Lori Jones transition occurring without public disclosure of where Or Peles landed, and Jones's prior Olive AI role inviting a question that a sophisticated buyer will surface during diligence. The DRG gap is structural: Nym's inpatient is professional-services only, so a $10B integrated delivery network looking for a single autonomous-coding vendor across the full inpatient stack will need a second tool. The architecture trade-off matters: rules-based CLU is strong on auditability but structurally less adaptive than deep-learning peers as clinical documentation drifts, and Nym has not published a rules-drift validation cadence.

Relative to the cohort: AKASA and Fathom are the closest direct competitors. AKASA is the broader RCM platform (coding + CDI + prebill + claim follow-up); Nym is coding-only with deeper specialty depth in ED and radiology. Fathom is the deep-learning specialist with stronger formal Epic posture (Toolbox + Blueprint), CVS Health Ventures backing, and a faster automation-rate trajectory (Your Health 95.5% in five months); Nym is the rules-based specialist with better named-customer density at the integrated-delivery-network scale and stronger KLAS scoring on implementation + support. CodaMetrix wins on inpatient DRG focus where Nym is structurally absent.

Revisit if: (1) Nym ships a HITRUST CSF Certification (any tier) and a public trust page; (2) Lori Jones's first 12 months produce a public commercial roadmap that addresses the DRG gap or definitively de-prioritizes it; (3) Nym publishes a per-specialty accuracy breakdown rather than an aggregate "over 95%"; (4) a peer ships a published rules-engine + audit-trail product that erodes Nym's transparency edge; or (5) Nym publishes a rules-drift validation cadence.

Right-of-reply gaps

• HITRUST CSF Certification roadmap - is "mapped to HiTrust" a stepping-stone to a real HITRUST e1 / i1 / r2 audit, or the endpoint?
• Public trust page roadmap - when does /trust or /security stop returning HTTP 404?
• Per-specialty accuracy breakdown - is "over 95%" the floor for every specialty (ED, radiology, outpatient surgery, outpatient visits, inpatient professional, urgent care) or the aggregate?
• DRG-coding roadmap - is professional-services-only the permanent inpatient stance, or is DRG on the multi-year roadmap?
• Or Peles current role post-April-2026 - board, advisor, departed?
• PHI data residency - does customer PHI ever leave US soil for Tel Aviv R&D operations?
• Customer PHI in rules-development loop - what is the standard DPA posture on using customer PHI to develop / refine the CLU rules engine?
• Foundation-model usage anywhere in the stack - is the CLU pipeline 100% rules, or are any LLM components involved (narrative summary, edge-case review, audit-trail generation)?
• Pricing model - per-chart, per-encounter, per-coder displaced, share-of-savings, or hybrid?
• athenaConnect marketplace listing - given athenahealth compatibility is claimed, why no marketplace presence?
• Native connector status for Cerner / Meditech / Allscripts - first-party connector or HL7 v2 only?
• Rules-drift validation cadence - how does Nym confirm the rules engine continues to map correctly as clinical documentation evolves?